Summary/abstract: Tuberculosis (TB) is the number one infectious disease killer in the world with 1.5 million deaths attributable to the disease in 2014. A critical knowledge gap exists given the lack of biological pathways and predictive biomarkers related to TB onset, progression and resolution. An improved diagnostic test for TB would have a major impact on global control of the disease and reduce mortality. Unfortunately, there are no validated biomarkers for TB onset, progression and resolution; the development of a TB biomarker is urgently needed. Current ultra-high-resolution metabolomics (HRM) methods have considerable promise for the development of TB-associated biomarkers. Several studies have identified distinguishing small molecule metabolic profiles in blood and other bio samples in individuals with active Mycobacterium tuberculosis (Mtb)-induced TB disease compared to uninfected controls, but no studies have explored whether metabolomics is predictive of TB outcomes, Mtb clearance from sputum. Further, all studies to date have used low-resolution metabolomics methods. Advances in mass spectrometer machine capability, coupled with recent advanced data extraction/analysis methods has considerably increased the dynamic range of metabolite detection by HRM in biologic samples, which now exceeds 20,000 species (>100,000 ions). Our HRM processing scheme provides capability to detect very low abundance metabolites, including Mtb cell envelope glycolipids, and explore regulation of human metabolic pathways. In a pilot study, we successfully identified 61 plasma metabolites that differentiated adults with active pulmonary TB from household contacts without TB. Differentiating species included specific Mtb-derived cell wall glycolipids and endogenous lipid mediator resolvins. Our recent pilot data shows that specific metabolites and human metabolic pathways, including those involved in drug, amino acid, and lipid metabolism, are associated with the propensity for sputum Mtb culture clearance over time. The ultimate goal of this exploratory proposal is to obtain novel data that may lead to the development of new TB biomarkers. We hypothesize that plasma HRM analysis can: 1) predict the propensity for successful anti-TB treatment by identifying Mtb-derived and endogenous metabolites (biomarkers) and human metabolic pathways (pathophysiology) associated with sputum Mtb clearance; and 2) successfully differentiate patients with active TB from controls without latent TB infection (LTBI) or TB disease. To test these hypotheses, we propose the following Specific Aims in this exploratory R21 project: Specific Aim 1: To determine whether plasma HRM analysis can identify Mtb-derived and endogenous metabolites that predict clearance of Mtb from sputum (change from positive to negative culture). Specific Aim 2: To determine whether plasma HRM analysis can differentiate adults with drug- sensitive or MDR-TB HIV co-infection from asymptomatic controls without LTBI.